H i C N Households in Conflict Network The Institute of Development Studies - at the University of Sussex - Falmer - Brighton - BN1 9RE www.hicn.org The Agency and Governance of Urban Battlefields: How Riots Alter Our Understanding of Adequate Urban Living1 2 Jaideep Gupte HiCN Working Paper 122 September 2012 Abstract: For the first time in close to 100 years, India reports higher population growth in its urbanised areas than across its vast rural landscape. However, a confluence of vast urbanisation and scarcity of resources has implied heightened levels of localised violence, centred in and around already impoverished neighbourhoods. This therefore has a disproportionate impact in further marginalising poor communities, and is at odds with the notion that cities are incontestably and inevitably the context of sustained poverty eradication. And yet, we know relatively little about the mechanics of security provisioning in Indian cities at large. The central argument in this paper is that violent urban spaces have a profound impact on how safety and security are understood by the state as well as the urban poor, thereby redefining the parameters of adequate urban living. I look in detail at how the 1992-1993 riots in Mumbai unfolded in a group of inner-city neighbourhoods, and find that specific acts of brutality and violence during the riots continue to shape current understandings of the „safe city‟. In doing so, I also find that the nature and form of informal urban space affects the mechanics by which the state endeavours to control violence, while individual acts of public violence function as markers that legitimate the use of, and reliance on, extralegal forms of security provision. 1 This research has been supported by DFID-ESRC Grant # RES-167-25-0481 “Agency and Governance in Contexts of Civil Conflict” and Microcon (www.microconflict.eu). Data collection in 2008, 2009 and 2010 was supported by the Harold Wingate Foundation and the Global Development Network Research Medal (Rule of Law). 2 Fellow of the Institute of Development Studies, UK. Contact: [email protected] 1. INTRODUCTION For the first time in close to 100 years, India reports higher population growth in its urbanised areas than across its vast rural landscape (Census of India 2011), with its urban population set to reach 600 million by 2031. And yet, India is only 30% urbanised, a rate much lower than China, Brazil or Mexico, leaving much room for further expansion. As in sub-Saharan Africa, only a minor proportion of India‘s urban growth is due to migration from rural areas as most of it occurs due to natural increases in the urban population as well as the reclassification of rural areas as urban (UNFPA 2011). By 2030, large cities could generate more than 70% of India‘s net new employment and produce 70% of its GDP. This evidence is consistent with a host of positive contributions urbanisation has had on development on the global scale. ―Almost all urbanisation among low- and middle-income nations is associated with economic growth. The world‘s largest cities are heavily concentrated in the world‘s largest economies. The more urbanised nations in Africa, Asia and Latin America generally have the highest life expectancies and lowest infant and child mortality rates. The nations with the worst health and living conditions among their urban populations are generally the least urbanised‖ (Satterthwaite and Mitlin forthcoming: 1). These positive contributions notwithstanding, we also know that this growth is built upon acute inequality. A third of the world‘s urban population live in slums, and the urban share of global poverty is increasing (Ravallion, Chen, and Sangraula 2007). In India, 37% of urban households live in accommodation of only one room or are homeless. Such a confluence of vast urbanisation and dense pockets of localised scarcity of resources often implies heightened levels of violence (Rodgers 2010, Tilly 1985). Recent evidence supports this claim: riots in India are persistent and widespread, with an average of over 64,000 riots per year over the last decade and 16 out of 28 states experiencing more than 1000 riots in 2010.3 Moreover, this violence can be located quite precisely, with a vast majority occurring in urban areas (Varshney 2002), and when large scale rioting breaks out in a particular city, not all neighbourhoods are prone to the violence (Gupte, Justino, and Tranchant 2012). It tends to be centred in and around impoverished neighbourhoods where ―insecurity has a disproportionate impact in further marginalizing poor communities‖ (United Nations Human Settlements Programme 2003: 183). That is, the debilitating effects of crime, conflict and insecurity are significant contributors to urban deprivation, while security also implies freedom from persecution and forced evictions, and 3 GoI. various years. Crime In India. New Delhi: National Crime Records Bureau, Ministry of Home Affairs, Government of India. provides for security of tenure. Violent urban spaces therefore have a disproportionate impact in further marginalising poor communities, and are at odds with the notion that cities are incontestably and inevitably the context of sustained poverty eradication. I argue that the lingering legacies of violent urban spaces have far reaching impacts, particularly for the urban poor, both over time as well as a result of spill-over into non-violent cities, by redefining the parameters of adequate urban living. This is most evident in terms of how safety and security are understood by the state as well as the urban poor, and as a result, how, and by whom they are provided. I attempt to answer how the presence of urban violence affects formal (i.e. state) conceptualisations of what a slum is on the one hand, and how informal spaces shape the state‘s efforts in controlling urban violence on the other, by looking specifically at the incidence of urban riots in Mumbai, India. I present a detailed description of how the 1992-1993 riots unfolded in a group of inner-city neighbourhoods, and find that current frameworks within which the state defines and relates to informal settlements fall short in accounting for the entire range of vulnerabilities and insecurities faced by the urban poor. Dataset and methodology The analysis is based on a dataset (Gupte 2011b) which includes two categories of data: 1) 41 in-depth qualitative interviews with men (58.5%) and women (41.5%) residents of Nagpada, Madanpura and Kamathipura – three inner city neighbourhoods in south-central Mumbai, conducted during three 5-month data collection efforts between February 2007 and December 2009. The sample is described in Table 1 below. Each respondent is identified in the paper only by his or her Respondent Code (RC). A further 20 declined to be part of the in-depth interviewing. Each interview lasted 120 minutes on average. A non-random purposive snowballing technique was used to recruit interviewees (c.f. Heckathorn 1997, 2002, Patton 1990, Salganik and Heckathorn 2004), which was initiated from 5 respondents (3 in Nagpada, 1 each in Madanpura and Kamathipura).4 Respondents were screened according to their residential status between 4 The initial 5 interviewees are not included in the dataset to maintain adequate degrees of separation between the December 1992 and January 1993, when Mumbai experienced its most intense bout of city-wide rioting, and only those who lived in, worked in, or were born into households that lived or worked in the case study neighbourhoods, were selected. Each interview was guided by three questions (alternatives in brackets) – ‗What is the meaning of security to you? (Who in your opinion needs security?)‘; ‗Who provides security (do you provide security) in this neighbourhood?‘; and ‗How is security provided?‘. Respondents were not specifically asked about the 1992-1993 Mumbai riots.5 Table 1: Description of the Sample Male Female Gender 58.5% 41.5% Average Median Oldest Youngest Age (years) 36.7 35.5 72 19 Illiterate/no schooling Up to primary school Graduate and above Education 25% 42.5% 32.5% Employment Unemployed Skilled Labour Unskilled Labour Student Housework only Status 22.5% 32.5% 20% 2.5% 22.5% Hindu Muslim Christian/Other Religion 37.5% 57.5% .05% Source: Adapted from Gupte 2011b interviewer and interviewees. 5 No monetary incentive was provided for partaking in the interview process. Some respondents, particularly those dependent on a day-wage, could only find time for the interview at night. In accordance with the ethical guidelines of the Central University Research Ethics Committee, University of Oxford, on research involving human participants, prospective respondents were first given a verbal explanation of the research project, and explained that the outcome of the research would result in academic research, and given the opportunity to opt out at any stage of the interview. Because their names had been suggested, in most cases the prospective respondents would make an effort themselves to make contact and fix an interview time. While each interview began by obtaining the informed verbal consent from the respondent, because no direct incentives were provided for participation, as well as the fact that the respondents came of their own accord to the interview site, their attendance was also seen as their consent to participation. 2) The dataset also includes 12 ‗elite‘ interviews conducted between 2007 and 2009 with police officers (7), inspectors (2) and constables (3) who were either posted to stations in the neighbourhoods of this study, or had direct knowledge of events during the 1992-1993 riots with specific reference to case study neighbourhoods. These two strands of data are brought together with a detailed exploration of the historical and current nature of urban form (layout, density, composition) of the case study neighbourhoods. The rest of this paper is organised as follows: in Section 2, I describe the current nature of the case study neighbourhoods in south central Mumbai, and explore the trajectory of the historical and social processes of segregation which have shaped the area into what it is today.